Re: A 1/4 wave Tesla coil

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "by way of Terry Fritz <twftesla-at-uswest-dot-net>" <FutureT-at-aol-dot-com>

>   
>  Secondary:
>  Skeleton form made of 12 polystyrene 1/2" rods and acrylic spacers
>  Size: 14" x 30"
>  Turns: 240
>  Wire: #16 stranded PVC covered machine tool wire, 1/8" in diameter
>  Wire length: 880'. This corresponds to about 280khz

Hi Skip,

Well this is quite a physically large coil.  In my work, I seem to be
seeing the effect that the size of the coil should match the input
power.  I also like the toroid to strongly dominate the secondary.
I would like to see a larger toroid, maybe 12" by 40" or so on that
secondary, but then maybe more power will be needed to drive
the system, but maybe not.  In any case, I would think that for
longest sparks, you would want a toroid size that gives just
one breakout or maybe two, even without a bump.  If you're
getting a lot of sparks all around the toroid without a bump,
then the toroid may be too small.  Low breakrate (120 bps),
seems to need larger toroids in general.  Sometimes for longest
sparks, the toroid has to be almost 1/2 the expected spark length.
I mean for a 105" spark, a 45" toroid may be needed at 120 bps,
but a 36" or 40" toroid may be OK.  Also, I like the minor toroid
diameter to be about 1/10th the spark length.  so for a 105"
spark, the toroid would be about 10"  to 12" thick.  Such a large
toroid may increase the spark length about 20% or so?  Of course
there's a range of sizes that will work well.   BTW, Alan Jones has
done excellent work with large toroids on his coil, and has gotten
something like 100" sparks from two of those photocopier type
transformers.  I've never heard of anyone else getting these kinds
of results.  I think Alan is using 1300 turns on his coil.  Alan is
also running his coil at 120 bps synchronous.  

>  Free air resonance: 400khz
>  Ctop: toroid 8" x 25" dryer duct with aluminum foil coating
>  Ctop capacitance: ~22pf
>  Secondary resonance with Ctop: ~280khz
>   
>  Primary:
>  Diameter: 19"
>  5 turns 3/8" copper tubing on 5/8" centers
>  Capacitor: .021uf
>  Primary resonance with cap: ~280khz

I like the primary coil to have a large surge impedance to keep
the gap losses low.  The wide diameter you're using helps to
keep the inductance somewhat high, but the coil may benefit
from having about 15 to 20 turns on the primary, and about
1500 turns on the secondary.  But this change may only gain
you about 10% to 15% spark length increase?  But it may help
even more than that.  Consider that I gained 10% just going
from 660 turns to 1600 turns on the secondary.  Also Richie B
gained 10% when he went from 1000 turns to 1500 turns.

Is the capacitor a low loss type?  If you're using Mylar, it's very
lossy, and will reduce the sparks by about 15% or so.  I guess 
mica would be OK, I know you use it sometimes.  I see you're 
using the matched size cap for the NST.  I would use a cap of
about 2.2X the resonant size, just to keep the voltage lower on
the NST to protect it.  I used to use the matched size, and I had
no problems, but the LTR size does seem safer.  I got the same
spark length either way.
>   
>  Primary/secondary system resonance: ~280khz 
>  System Q: over 150 but measurement limited by my instrumentation
>  K: Not measured at this time but expected to be .21 or higher
>  Best coupling is obtained with top of primary about 1/2" above bottom of
>  secondary

I don't think the Q is very important on a spark gap coil.  The Q tends
to be important on a CW or tube coil because the coil is pumped up
resonantly.  The 1/4 wavelength for the wire length is not important
either.  That's a concept that has been disproved.
>   
>  Power: 15/120 NEST
>  Break: SRSG, 1800 rpm and 120 BPS
>  Contact spacing: 0.2" total, so far but still working up to larger spacing

By contact spacing, you mean the gap spacing between electrodes?
I usually keep this quite narrow.  If it's too wide, it can cause intermittent
operation, especially at low powers.  Regarding the gap phasing.  In some
work I did, I found that there were other places where I could set the phase,
which worked, but gave a terrible power factor.  I like to monitor the
wattage with a wattmeter, and the current with an ammeter, and select
the gap phasing which gives the best power factor, and use a PFC cap.
(I may be becoming confused with work I did using a PT to power the
coil, I'd have to check my notes.)
>   
>  Performance:
>  Operating frequency: 280khz
>  Spark length with no breakout tab: multiple arcs 48"(+)
>  Spark length with breakout tab: single arcs 60"(+), very, very thick
>  Interesting note: With no Cop, arcs are close to 42" from top terminal, but
>  coil is obviously poorly tuned and much corona around top turns of 
secondary.
> 
>   
>  This coil operates at the design frequency. There is no indication that 
any 
> ion
>  cloud adds capacitance to the system that might lower the operating 
> frequency. 
>  This coil was designed basically using Medhurst and Wheeler formulae. No 
> tuning
>  was required after assembly except to phase the SRSG and to increase the
>  coupling between primary and secondary.
>   
>  Changing the primary turns only reduces the output length of sparks. 
> Question
>  to those on the list with experience: The output sparks are very thick and 
> fat
>  but only reach 60"(+). The coupling is as close as possible, just before 
> racing
>  sparks start, and changing the SRGS phasing only decreases the length. I 
> would
>  appreciate any suggestions to increase the spark length. John F., I know 
you
>  have done much in this area and I would appreciate your comments in 
> particular.
>  Note also that the above indicated operating characteristics occur with 
> about
>  100 volts input and additional voltage up to full line voltage seems to 
make
>  little if any difference in spark length.

I don't understand why increasing the voltage doesn't improve the
spark length.  Usually I see a nice spark length increase at that
power level. I assume you adjust the gap phase at full power, and
verify the tuning of the primary at full power.  Often at full power, the
coils needs to be tapped out one more turn or so (maybe a half
a turn).  You have to tune the coil so it barely works at low power,
but then the sparks break out with a vengence at high power.  You
may have to turn up the power quite a lot before the sparks break
out.  I see this all as evidence that ion cloud effects are lowering
the resonant frequency.  The coil may be working itself into tune
as the sparks grow.  That NST can draw about 3600 watts or so with a resonant
sized cap, or with an LTR cap if you raise the voltage to 140 volts
input.  My 12/30 NST draws about 720 watts (double the rating),
when it's run at 120V with a matched size cap.  When I use an
LTR cap, I have to raise the voltage to 140V to get 620 watts
power input for the same spark length.  However, if I rephase my
gap, I can cause the resonant sized cap to require 140V input for 
best spark length.  This gap setting is more efficient and draws 
only 620 watts.

If we assume that your NST can draw 3600 watts with a matched
size cap, at 120V input, then the coil should give about 105"
sparks if everything is optimized I would think.

It's an interesting project, and I'll be interested to hear how it
progresses.

John Freau
>   
>  Skip Greiner
>